VARIATION OF SALMONELLA PULLORUM 35 



materially in appearance from the uninoculated. In some instances the contents, 

 after an incubation period of two weeks, resembled the interior of diseased ova in 

 adults and unabsorbed yolk in young stock. 



The observation that variants may pass through the egg unchanged in their 

 colony form is substantiated by the above findings. 



Behavior of Variants in the Natural Animal Host 



In this experiment the behavior of the variants in chicks and adult fowl was 

 studied. Special attention was directed toward determining the stability of the 

 types in the living host. Pathogenic and serologic characters were studied in 

 greater detail in another series of experiments. Pullorum disease-free stock, 

 including chicks and mature fowl, was employed in this experiment. Both variants 

 and typical normal strains were investigated, the latter being regarded chiefly 

 as controls. Saline suspensions were prepared from 24-hour agar cultures. The 

 concentration was adjusted to a turbidity of tube 2 of McFarland's nephelometer 

 scale in all instances except one, in which the turbidity was equal to tube 1.25. 



Four different lots of chicks were fed or inoculated with suspensions of the 

 organism. The chicks exposed to the different strains were confined in separate 

 cages and cared for in such a manner as to minimize the possible hazard of cross 

 infection. In a few instances the chicks were intentionally chilled in order to 

 reduce their resistance to the invasive power of the organism. A period of approx- 

 imately two weeks elapsed between the time of inoculation or feeding and the 

 necropsy. Chicks that succumbed during the observation period were necropsied 

 within a short time after death. Cultures were made from the liver, spleen, yolk, 

 and in some instances other organs or tissues, depending upon the pathological 

 picture encountered. Tissue impression smears were made from chicks in the 

 first few inoculated series for the purpose of determining the presence and morphol- 

 ogy of the organism in the host. Only those smears prepared from tissues which 

 yielded S. pullorum in pure culture were stained with a dilute solution of Hucker's 

 gentian violet, and examined. The isolated cultures were checked for type in 

 broth containing dextrose, maltose, lactose, sucrose, and in some instances dulcitol ; 

 by the Gram staining method; also in a number of cases by the agglutination test. 

 No. 1. In this lot, 26 chicks were fed with three variants (C, D-l, and E) 

 and one typical strain designated 11,706, which had been recently isolated from 

 the ovary of a hen. The dose was 0.03 cc. The results obtained from feeding 

 the different organisms are presented in Table 9, and show that the percentage 

 of infected chicks was very low. Three of the four chicks which revealed S. 

 pullorum had been inoculated with the typical strain, 1 1,706. The recovered organ- 

 ism resembled the original strain in colonial and cellular morphology. The strain 

 recovered from the group infected with Variant E also appeared identical with 

 the original, and was agglutinated by positive pullorum serum. The cells settled 

 out in the presence of negative serum and in the antigen control. A few of the 

 cultures were isolated in duplicate on meat extract and liver infusion agars. 

 Infrequently growth occurred on one medium and not on the other, especially 

 when but few colonies were isolated from an organ. Occasionally only one colony 

 was observed on a plate, which suggests that the amount of tissue used for inocu- 

 lating is an important factor in obtaining positive isolations. 



No. 2. Eighty-two chicks obtained from the same source as those in No. 1 

 were fed and inoculated with six variants (B, C, D-l, E, Vl-a, and IX-b) and 

 four so-called smooth types (IV, XIV, 11,706, and 18,292). For a description of 



